Average point margin throughout the game. Solid line = mean, shaded band = 25th-75th percentile range, dashed lines = min/max. Filter by period or opponent quality.
Win probability at each margin and minute. Green = high win probability, red = low.
Positive thresholds: games that reached that value or better. Negative thresholds: games that fell to that value or worse.
Average margin at key points. Click headers to sort.
| Team | 6 min | 12 min (Q1) | 18 min | 24 min (Half) | 30 min | 36 min (Q3) | 42 min | 48 min | Final |
|---|
| Team | Comeback Wins | Avg Deficit | Max Deficit | Wins No Trail | Blown Losses | Avg Lead | Max Lead | Losses No Lead |
|---|
| Team | Frequency % | Total | In Wins | In Losses | Win % | Successful | Failed | Success % | Avg Start | Avg Duration | Avg Diff |
|---|
| Team | 1min Best | 1min Worst | 3min Best | 3min Worst | 6min Best | 6min Worst | Quarter Best | Quarter Worst | Half Best | Half Worst |
|---|
| Team | 1min Best | 1min Worst | 3min Best | 3min Worst | 6min Best | 6min Worst | Quarter Best | Quarter Worst | Half Best | Half Worst |
|---|
Non-overlapping scoring runs per game. Bottom-right = elite (explosive but composed).
Average point differential around timeouts (-2min to +2min). Normalized: 0 = margin at timeout moment.
On-court percentage over game time. Select a team, then tick players.
No players selected
Games clustered by margin progression (from winner's perspective). Hover for details.
Season 2025-26 | Showing 0 of 0 games
$L$ = number of lead changes in the period
$m(t)$ = point margin at time $t$ (home − away)
$\tilde{t} = t / T$ = normalized time (0 at period start, 1 at period end)
$\mathbf{1}\{\cdot\}$ = indicator function (1 if true, 0 otherwise)
| Parameter | Symbol | Value | Description |
|---|---|---|---|
| CLUTCH_ALPHA | $\alpha$ | 0.3 | Possession bonus when trailing team has the ball |
| CLUTCH_BETA | $\varepsilon$ | 1.0 | Denominator offset (smooths small margins, prevents ÷0) |
| CLUTCH_GAMMA | $\gamma$ | 0.15 | Lead change multiplier |
At each moment, tension is inversely proportional to the margin:
Examples:
Later moments matter more via quadratic weighting:
For Q4 ($T = 720$ seconds):
More lead changes = more exciting game:
Example: 4 lead changes → $1 + 0.15 \times 4 = 1.60$
Each overtime period contributes at 50% weight:
| Clutch Index | Interpretation |
|---|---|
| < 0.05 | Blowout — game decided early |
| 0.05 – 0.15 | Comfortable win — some tension |
| 0.15 – 0.25 | Competitive — close game |
| 0.25 – 0.40 | Clutch — exciting finish |
| > 0.40 | Instant classic — overtime thriller |
Velocity (pts/min) vs Current Margin | Forward window: 60s